LevelAI has a fantastic structure for analyzing our calls, the idea is amazing. However, the UI needs improvements so we can see more benefits on using it and have a better user experience. Review collected by and hosted on G2.com.
As an agent, it is not clear what the standards are for analyzing the call sentiment, and not necessarily a call that has been classified as negative was actually a bad one, so the numbers seem to be always off. The dashboard has too much information, it is easy to get lost and not know where to go for simple tasks. Additionally, it would be great to see our case numbers easily, so we could to Salesforce and actually see the case details. Review collected by and hosted on G2.com.
78 out of 79 Total Reviews for Level AI
Overall Review Sentiment for Level AI
Log in to view review sentiment.
- It captures data in various areas, allowing for thorough information to be pulled quickly
- The ability to personalize your own dashboard with graphs and other information specfic to your role is fantastic. This allows you to create permanent visuals and reports for various areas of the business, allowing for increased efficiency each time that information is needed.
- It will analyze data for any time periods and is very user-friendly
- We are able to align our quality standards across teams and partners in a large organization.
- The transcripts are easy to navigate, and clicking on various parts will take you immediately to that part of the communication.
- Tech issues are handled quickly, their customer services is wondeful Review collected by and hosted on G2.com.
- It is delayed with picking up the start of phone calls, so it doesn't capture the first minute of conversation on inbound calls. Due to the delay, it consistently grades incorrectly for the first couple of rubric questions.
- There is also a significant delay with data filtering into Level AI
- It would be nice to allow calibrations to have more than one moderator, allowing for calibration across departments. Right now you can only have one moderator, and the "correct answers" don't stick once you leave the calibration.
- During calibrations, the moderator will share their screen so participants can view their answers, but the screen is crowded and makes it difficult to see for those working remotely.
- If anyone has even a minor accent, the transcript is incorrect the majority of the time. Review collected by and hosted on G2.com.
Outstanding transcription accuracy, easy to use platform, screen cap functionality, scoring accuracy is very accurate and allows for all recorded calls to be scored, easy to identify concerns based on keyword search ability, terrific reporting capabilities. Review collected by and hosted on G2.com.
Certain things difficult to capture and score accurately pertaining to multiple topics in one sentence. Review collected by and hosted on G2.com.
The tool is now being used for quality evaluations within my organization which make contacts easier to evaluate and allows frequent use without issues. It is very easy to use.
Level AI helps to capture real time interactions across different channels such as email, chat and phone support. It is most times accurate and very efficient! It gives you the opportunity to add detailed feedback to support the growth of agents. It is a great QA tool and I like the concept. Never had the need to contact customer support. Review collected by and hosted on G2.com.
While Level AI is efficient, there are a few areas that could be improved. The notification of an evaluation seems to be delayed.
The layout can be revised to make it easier to find information as it feels overcrowded and may cause confusion.
I also noticed that scoring isn't always accurate due to language barrier within the system which results in an inaccurate score. Review collected by and hosted on G2.com.
Apart from the fantastic aesthetics of the platform itself, my favorite thing about Level is the customization capabilities. The ability to slice and dice reports by any piece of metadata that we import has been extremely helpful in gaining insights into agent and call center performance. The Help Center is also incredibly robust and a great resource. Review collected by and hosted on G2.com.
The lines of communication between my company and our POCs at Level could be more open and clearly defined. Additionally, while I think it will improve throughout our relationship with Level AI, the platform itself seems more built out for customer support call centers which is not necessarily aligned for my company's particular use case. Review collected by and hosted on G2.com.
The layout of the page makes everything easy to find. My primary use of Level AI is for QA on my interactions with customers. The auto QA is really helpful as it pinpoints what was/wasn't done based on the different types of information required to meet the goal. For example, I am required to first validate the account (get the customer's name) before I provide any information to the customer. The auto QA would state, once I fully achieve this objective 'The agent addressed the customer by their name at the beginning of the conversation, which is a good practice to make the customer feel valued and recognized.' It's very straight forward and I love how it adds the little tip to indicate why you should take certain actions and how it improves the customer experience. During weekly coaching sessions my supervisor and I'll pick random contacts to review and the section for coaching allows you to keep a track of which interactions were evaluated and the analytics area shows your progress. It also shows were you fell back so it allows me to better manage my QA targets and easily figure out what I didn't do for the few that I didn't meet the goal. Review collected by and hosted on G2.com.
The only issue I have with the tool is that it doesn't seem to be able to distinguish the greeting in an email. It always shows Hi {person} instead of the customer name which makes the auto QA state that I didn't verify the account. Review collected by and hosted on G2.com.
The transcription is the only portion of the platform that makes my job faster as an auditor, since I no longer need to listen to the entire call myself and can scan through the transcript for relevant information and only spot-check the audio when needed. Review collected by and hosted on G2.com.
Call ingestion is delayed by 24 hours or more, so we cannot monitor same day calls reliably. When the calls do ingest, many of them port over without audio so we have to trust the transcript is 100% as we have no way to double-check by spot-listening to the call itself. The "AI learning" is nowhere near being ready to understand the context and nuances necessary to be reliable in its answer choices, so many of its "answers" are deleted and replaced by our team manually, which is the exact opposite of what we hoped to do. Conversations with the Level team seem to go in circles without requested updates or fixes being added unless the problem is a complete work-stopper for our team. Review collected by and hosted on G2.com.
This generative AI tool has completely transformed the way our organization handles scoring across all communication channels—calls, chats, and emails. The accuracy is nothing short of impressive, consistently hitting near 100% across our custom scorecards. It’s like having a super-efficient team member who never misses a detail!
What I love most is the seamless integration into our existing workflows. It saves us countless hours of manual scoring, allowing our team to focus on what really matters—providing exceptional service to our customers. Plus, the insights generated help us identify areas for improvement and celebrate our wins. Review collected by and hosted on G2.com.
While Level AI is impressive, there are a few areas where it could improve. Occasionally, the AI might struggle with nuanced conversations or context-specific scoring, which can lead to minor inaccuracies. Review collected by and hosted on G2.com.
The standout feature of Level AI is its exceptional customer service! Their team takes the time to walk you through the process, ensuring you feel supported every step of the way. They also hold weekly meetings until your team feels confident. Plus, they customize the platform to fit your company’s unique needs. Communication is effortless, and the team is genuinely easy to talk to. Review collected by and hosted on G2.com.
Like any AI platform, training the system to recognize specific patterns can be time-consuming, but it’s an essential step for achieving optimal results. This process is understandable given the complexity of tailoring the AI to meet your needs Review collected by and hosted on G2.com.
The platform’s reporting tools are a standout feature, offering in-depth insights into customer sentiment, common issues, and agent performance. This has been crucial for improving both agent training and overall service delivery. Review collected by and hosted on G2.com.
One aspect of Level AI that I think could be improved is the lack of real-time notifications when an audit is completed for my advisors. Currently, we have to manually check for updates, which slows down our ability to provide timely coaching and feedback. Having instant notifications when an audit is done would significantly enhance our ability to act quickly and support our team’s development Review collected by and hosted on G2.com.
1. Auto scoring can be used as a guide to manual scoring. This can be a time saver in cases where multiple samples need to be reviewed in a short time frame.
2. I like the filter feature which helps to give view of results within a set timeframe, for selected specialists as well as seleceted evaluators Review collected by and hosted on G2.com.
1. The pending score is not shown while the evaluation is being done. The score only appears after the initial submission
2. The description of what is required for each parameter no longer shows, which was a feature at the initial launch. This was handy when reference need to be made to the guideline for the scoring
3. There is some delay with the evaluations being captured in the filter count after it is done. Sometimes it takes over 10 minutes it seems. You may complete an evaluation and the score is shown and you're sure that it was done. When you visit the filter to check on evaluations comoleted by you, they take a few minutes to reflect on the report.
4. Persons have reported that they see where the monitoring tool is still monitoring after a contact has ended
5. Comments added are shown as 'grey' and need to be highlighted for easy reading. A bright 'black' color would make it easier to read.
6. There are instances in which Level Ai captures and reviews an email responce which was a follow-up to an intitial conversation and not the conversation itself Review collected by and hosted on G2.com.
